Philip Metzger

As mentioned in Discord: https://discord.com/channels/968932220549103686/969829516539228222/1047922968543637504 I'd like to implement a `jj run` command which would allow integration with https://pre-commit.com, it should come with the usual set of options and should...

I like the idea as it goes in the direction which @arxanas proposed here https://github.com/martinvonz/jj/discussions/1905#discussioncomment-6539851. Ideally this would also be implemented in [scm-record](../../../arxanas/scm-record) itself.

I broadly support this FR. I only take issue with the thing below. > ## Design nits > The two most important nits I can think of are the following:...

> don't follow the whole discussion, but do we need some mechanism to manipulate existing description by template language (like sed or awk)? I hope not Agreed, the suggestion with...

Sorry to barge into your conversation, we recently talked about renaming branches to bookmarks like Mercurial, which definitely fits better to the current `jj branch` implementation. We then should implement...

And since I'm here anyway. > Here are some characteristics I'd like the solution to have: > > * Rather than a new command, this should be a flag on...

Sorry, the above took a bit longer than expected. > that sounds good. But would that make `--advance-branches` unnecessary? Yes, it would make `--advance-branches` unnecessary, but until something like it...

> I think we have pretty different workflows and/or philosophies, but I do want to understand each of your points and where you're coming from. I generally don't think we...

Here's the last part: > > This is a straight "No" from me. Maybe it has a place in your downstream if you find the time. > > I sort...

I'm slowly inching closer to the implementation of `run`, as most of it is in the 3rd and 4th checkmarks of #1869. But atm you'll still need to run the...